| Age | Commit message (Collapse) | Author |
|---|
|
|
|
The 3D destination shares the same cache so we don't have any trouble with
the later commands needing the writes flushed inside of the same batchbuffer.
|
|
|
|
|
|
Instructions such as RCP, RSQ, LOG must smear the result of the function
across the dest register's X, Y, Z and W channels (subject to write masking).
Before this change, only the X component was getting written.
Among other things, this fixes cube map texture sampling in GLSL shaders
(since cube lookups involve normalizing the texcoord).
|
|
|
|
And check opcode number to avoid crashing on driver-private opcodes.
|
|
front_region may be null.
|
|
Plus fix up a debug printf.
|
|
The i965 hardware cannot do GL_CLAMP behavior on textures; an earlier
commit forced a software fallback if strict conformance was required
(i.e. the INTEL_STRICT_CONFORMANCE environment variable was set) and
2D textures were used, but it was somewhat flawed - it could trigger
the software fallback even if 2D textures weren't enabled, as long
as one texture unit was enabled.
This fixes that, and adds software fallback for GL_CLAMP behavior with
1D and 3D textures.
It also adds support for a particular setting of the INTEL_STRICT_CONFORMANCE
environment variable, which forces software fallbacks to be taken *all*
the time. This is helpful with debugging. The value is:
export INTEL_STRICT_CONFORMANCE=2
|
|
|
|
|
|
|
|
|
|
Thanks to Brian for the refactor commit
|
|
|
|
|
|
|
|
|
|
|
|
Move _mesa_init_all_x86_transform_asm() into x86.c so that common_x86.c
has no dependencies on the vertex transformation code.
Plus some comments and clean-ups.
|
|
Namelly, FlushMappedBufferRange takes a subrange relative to the original
range.
|
|
|
|
|
|
|
|
|
|
This fixed a minor bug but broke everything else.
This reverts commit 579ef8ab1984d895867f547afa60b3bec4c4599a.
|
|
|
|
|
|
Add asserts for expected values on wakeup and flush. Remove cases where
this flag is set or cleared except when waking up and flushing vbo module.
|
|
See bug 20319.
|
|
|
|
|
|
The core reference counting code is centralized in p_refcnt.h.
This has some consequences related to struct pipe_buffer:
* The screen member of struct pipe_buffer must be initialized, or
pipe_buffer_reference() will crash trying to destroy a buffer with reference
count 0. u_simple_screen takes care of this, but I may have missed some of
the drivers not using it.
* Except for rare exceptions deep in winsys code, buffers must always be
allocated via pipe_buffer_create() or via screen->*buffer_create() rather
than via winsys->*buffer_create().
|
|
For some reason makedepend doesn't seem to find the dependency on this
header in these two files. Directly including the header is a work-around.
|
|
Not needed since renderbuffers are ref-counted.
Fixes progs/demos/shadowtex.
|
|
|
|
When a hw driver fell back to swrast, swrast wasn't always getting informed
of program changes. When fixed function is translated into shaders, flags
like _NEW_LIGHT, _NEW_TEXTURE, etc. should really signal _NEW_PROGRAM.
In this case, swrast wasn't seeing _NEW_PROGRAM when new fragment shaders
were generated.
|
|
|
|
Conflicts:
scons/gallium.py
src/gallium/auxiliary/pipebuffer/pb_buffer_fenced.c
src/gallium/include/pipe/p_defines.h
src/mesa/vbo/vbo_exec_api.c
src/mesa/vbo/vbo_exec_draw.c
|
|
|
|
Using PIPE_BUFFER_USAGE_DONTBLOCK.
|
|
Previously would have to allocate a new VBO after firing a draw command
as subsequent call to Map() on old VBO might block if the driver had
submitted the commands to hardware.
|
|
Will be needed in coming GL extensions (GL_map_buffer_range, GL 3.0).
Will be used by the vbo module to avoid reallocating vbo's at each
draw primitive call.
|
|
Provides notification to the VBO modules prior to the first immediate call.
Pairs with FlushVertices()
|
|
|
|
|
|
drawing/reading
It seems that XGetImage() from a depth 32 TrueColor window is flakey.
Drawing with XPutImage() instead of XPutPixel() seems to work better, but
still not perfectly.
Keep using the original code for now until more is learned.
|
|
Another conditional can be avoided.
|
|
Need to clear the _ReallyEnabled field before possibly continuing the loop.
Also, set _Current pointer to NULL if the unit is no longer enabled.
Fixes piglit lodbias regression
|
